After taking 10 x 5 minute Luminance images of NGC246 on 21st Oct, I then managed to grab a set of RGB images a few nights later, on the 23rd October. R & G images were 15 min exposures with B at 20 mins.
This is my 2nd LRGB and the process is becoming a little clearer, and less daunting, as I come to grips with the techniques involved. I was fortunate enough to have Jase’s working file (from his reprocess of my M16 image) to use as a template. This was very useful in guiding me through the process for the various Layers in Photoshop CS3.
The (earlier) L and (later) RGB frames were not quite aligned, so I had to crop the final image to remove the mis-registered region. I applied some mild Maximum Entropy deconvolution which on reflection, may have been a little overdone. Further work is also required to minimise colour casts and achieve a better colour balance, but I’m happy with what I’ve done so far.
C9.25, F6.3 Reducer/Corrector and SBIG ST7E CCD camera.
Thanks for looking.
Cheers
Dennis
PS – Scott Alder (Tornado33) has a nice wide field of this PN on Page 72 in the Nov/Dec 2008 edition of AST.
Last edited by Dennis; 03-11-2008 at 05:41 PM.
Reason: Changed Richardson-Lucy to Maximum Entropy
Nice work Dennis. A significant improvement to your last. I don't think you've overdone the deconvolution. If anything, I think you can tighten up the details and stars even further. Perhaps try a few more iterations to confirm the theory and validate how much the data will accept before artifacts arise. Given the over sampling of your imaging rig, deconv can work wonders on the data. Colours look good, so you balance is on the money. Re: decon, try a few different strengths and save them as separate images. You can then blend each image within PS providing greater control of the detail extraction and avoid any halos around the stars. Look forward to seeing more.
Beautiful capture Dennis.
The colours are lovely.
What do you use to apply the Richardson-Lucy deconvolution?
Is there a plugin for PS or do you use another app?
A nice deep image there, I like the inner detail youve recorded.
Scott
Thanks Scott – I’ll have to chase down the missing green that appears on your terrific image in AST, along with the (relatively) few examples that I’ve seen on the web. I only managed 1 sub each of RGB, so maybe more colour data would help?
Nice work Dennis. A significant improvement to your last. I don't think you've overdone the deconvolution. If anything, I think you can tighten up the details and stars even further. Perhaps try a few more iterations to confirm the theory and validate how much the data will accept before artifacts arise. Given the over sampling of your imaging rig, deconv can work wonders on the data. Colours look good, so you balance is on the money. Re: decon, try a few different strengths and save them as separate images. You can then blend each image within PS providing greater control of the detail extraction and avoid any halos around the stars. Look forward to seeing more.
Thanks Jase – it has been a real help having the M16 LRGB example you did for me, it made the penny drop! Previously I had been thrashing around, making ad-hoc adjustments and your M16 re-process has provided me with a defined approach which I can now repeat.
I was a little unsure about the de-conv and found that Max Ent gave a better result than RL, so I’ll certainly have more of a play around. Now that I have a better understanding of layers, the whole process has become so much clearer; and more importantly, fun!
Beautiful capture Dennis.
The colours are lovely.
What do you use to apply the Richardson-Lucy deconvolution?
Is there a plugin for PS or do you use another app?
Hi Andrew
Thank you for those nice words of encouragement. I should have written Maximum Entropy and not RL, as Max Ent was the deconvolution process I used as if gave me slightly better results than RL. Maybe it was just the parameters I chose?
I also have a few deconv procedures in applications such as ImagesPlus and AIP4Win, but haven’t had much success with them.
I purchased the Astraimage Plug Ins for CS3 which include MaxEnt and RL (plus a few other procedures) and this is the 1st time I have used it, on NGC246. The (Astraimage) DDP filter doesn’t seem to do much for me, compared to the equivalent routines in ImagesPlus and AIP4Win which seem to have worked better in the few examples I have tried.
Great image and colour. I have never used and don't really understand deconvolution but maybe it'stime to learn.
Thanks Matt!
From my very limited understanding, I think that deconvolution works best when the image is oversampled and you have a good signal to noise ratio, so the real data hidden in there can be teased out.
The danger is that a poor S/N ratio will result in “artificial” details appearing in the image, where noise is mistaken for “real data”.
looks good,
I love the tiny targets that are a real challenge to image!
well done indeed
frank
Thanks Frank!
I loved your Thor’s Helmet and now that I am feeling a little more confident and skilled with LRGB processing, that will be my next project, although only at 1400mm and not at your 3900mm!
